In a color diffusion transfer process, a color image is formed by utilizing a difference in diffusibility between a dye image-forming compound and a dye compound (which forms a color image in an image-receiving layer) derived from the dye image-forming compound. Typical examples of the dye image-forming compounds include dye developers, dye-releasing couplers (DDR couplers), and dye-releasing redox compounds (DRR compounds) as described, for example, in Photographic Science and Engineering, Vol. 20, No. 4, pages 155 to 164 (July/August, 1976) and T. H. James, The Theory of The Photographic Process, Fourth Edition, pages 366 to 372, Macmillan Publishing Co. Inc., New York (1977). Yellow dye-releasing redox compounds include azo-pyrazolone type yellow dye developers and yellow dye-releasing redox compounds having an aryl group at the 1-position of the pyrazolone nucleus, as described, for example, in U.S. Pat. Nos. 3,309,199 and 4,013,633 and U.S. Published Application (USB) No. 351,673. However, the dyes formed from the compounds specifically described in these patents have certain technical problems, including a low transfer speed and a low maximum transferred image density (D.sub.max). Although improved yellow dye-releasing redox compounds are described in Japanese Patent Application (OPI) No. 111344/79, further improvements have been desired. (The term "OPI" as used herein refers to a "published unexamined Japanese patent application".)